Canonical statement
View source LaTeX
There is a language in \(NP\) that is not decided by any
family \((C_n)_{n\geq1}\) of Boolean circuits of size polynomial in \(n\),
where \(C_n\) decides membership for all inputs of length \(n\); in
symbols,
\[
NP\nsubseteq P/poly.
\]
Here \(NP\) is nondeterministic polynomial time, and \(P/poly\) is the
class defined by such polynomial-size, nonuniform circuit families.Notes
The conjecture asserts that some language in cannot be decided by any family of polynomial-size Boolean circuits, in symbols . Since , this is a nonuniform strengthening of . Its modern formulation dates from around 1980, when Karp and Lipton studied the consequences of the opposite inclusion [KarpLipton1980].
Karp and Lipton showed that would collapse the polynomial hierarchy [KarpLipton1980], which is the main structural evidence for the conjecture. Yet unconditional lower bounds remain scarce: no superpolynomial circuit lower bound is known for any explicit language in , and the strongest unconditional results concern restricted circuit models, such as Williams's lower bound against nonuniform circuits [Williams2014Circuit]. The general theory is developed in Arora and Barak [AroraBarak2009].
The conjecture is open: proving it requires a superpolynomial lower bound against general Boolean circuits for an explicit problem, a barrier that has resisted all known techniques.
References (3)
- [KarpLipton1980]
Some connections between nonuniform and uniform complexity classes
Open ↗Richard M. Karp and Richard J. Lipton · 1980 · misc
- [AroraBarak2009]
Computational Complexity: A Modern Approach
Open ↗Sanjeev Arora and Boaz Barak · 2009 · misc
- [Williams2014Circuit]
Nonuniform ACC circuit lower bounds
Open ↗Ryan Williams · 2014 · misc
The boxed statement is the canonical open formulation — not a stronger variant or a related research program. The status reflects the catalog's last review; do your own literature search before investing serious effort.